AlgoStrats Blog

  • By Marco Mayer
  • Posted Apr 18, 2017

Are there sure thing edges in trading?

Thank you for your feedback on my last article “Having an edge in the markets”. Some of you have been brave enough to take that article to the next logical step and ask the question most traders hardly dare to ask.

Here’s an email I got: “I have experienced all the things you described. The thing that I always wonder about is how do you know if you have a real trading edge? Every time I have a winning period and then move into a period of drawdown I start asking myself, “Do I really have an edge or am I now just giving everything back because I don’t?”. Is there a way to have an assurance that your methodology has an edge? Is it possible to have that confidence?”

Definitely, one of the hard questions when it comes to trading. And one that makes most traders feel very uncomfortable for a reason. I myself kind of blanked out each time it came to my mind for years. Why? Probably because I already knew that I wouldn’t like the answer too much. And it’s much easier to just move on with some magical thinking and decide to “trust in your guru” or “believe what you’ve read in a book”.

But one day every honest trader has to face this, so let’s not lose time and dive right into it.

First, let’s redefine the question. I’ll assume you have read the last article and if you didn’t, now is a good time to do so. In the article, we discussed that a casino has an edge at the roulette table, a real edge that is indisputable and 100% real. You can explain it in a few sentences, it has been there 10 years ago and will still be there in 10 years as long as the rules of the game don’t change. The house has an advantage, and the owners of the casino can be 100% confident about this.

How’s the situation in trading? Are there edges equal to the house advantage of a casino? And the answer is actually yes, here are a couple of examples:

  • When most trading still happened in the pits, traders who had a seat at the exchange and were trading directly from the pit had an edge. They could see the big players placing their orders, immediately see/hear/feel when the market started to move and jump on the move early. By the time the retail traders called in their orders to join the trend they could already cash in. Notice that almost all of these successful pit traders failed to make money when they were forced to trade in front of a computer screen without that locational advantage. Many of them are now trading gurus…

  • Market makers who were front-running orders. These are almost gone these days in electronic trading, but HFT trading algorithms have actually taken their place. They usually have a very small speed advantage, being a bit closer to the exchanges than everyone else. And a huge part of their game is running the same strategies as the human market makers did back in the days. With smaller amounts but a lot faster and at much higher frequency.

  • Arbitrage opportunities, trading the same stock but on different exchanges for example to exploit price differences. This was a really nice edge but the margins got smaller and smaller over the years as the markets became more efficient. Nowadays there still are some opportunities in markets like cryptocurrencies or developing markets in countries that are just opening up to the global world. But they’re not easy to find and the profit margins usually are very small.

There’s more, but the truth is that all of that kind of sure thing edges I know of are gone or too expensive to exploit for a private trader. One example of this is HFT trading where you need a serious amount of money to get a fast enough connection to the exchanges and super cheap trading costs to make this really work. Unless you have that, HFT trading is a sure way to ruin due to the trading costs you have to pay and getting beaten constantly by faster HFT trading algorithms.

Now let’s move on to the kind of edges most of us are looking for on a daily basis. These aren’t sure thing edges like the one the casino has. Let’s call them speculative edges. One might be that during a certain time of the month’s stocks tend to move higher. Or a chart pattern that increases the odds of prices to move in a certain direction. Or a trader who believes he has exceptional gut instincts about where the markets will move. It could also be a specific system/method like the well-known turtle system.

What differentiates all of these speculative edges from the ones we discussed before is that they’re always uncertain. A trading method like the turtle system might work for a decade and then stop working. The reason for this might be that too many people are doing the same thing or because the markets change. The same can be said of a chart pattern. Markets, market participants and market relationships change all the time. Some markets even completely disappear. And so do market inefficiencies.

Because of that, truth is you can never be sure if what you believe to be an edge actually still is an edge.

But there are things you can do to increase the odds of an edge being valid and to make this a bit less of an issue.

First of all, don’t just blindly believe and bet your money on a chart pattern/method/idea because it’s on some website, a trading guru tells you about it or you find it in a magazine/book. I can honestly say that I’ve looked at most of what’s out there in the trading world (for free and expensive/premium products). Truth is that even when you ignore the obvious bullshit it’s very hard to find any real edges in what’s on offer out there.

Also just because something sounds logical and reasonable doesn’t mean it will actually give you an edge. This is a very common scheme in the trading education world. We all love these little stories and explanations as for other areas of life, this actually works quite well. But not in trading.

Here’s an example: “this price pattern shows you clearly that the big players in this market bought at price X, after which prices moved higher very strongly. So the bulls clearly won and the traders who are short have to cover their position once price Y is hit.” Sounds logical, makes some sense, could be true right? But all of this is just an interpretation of what everyone can see on a price chart. It doesn’t mean that bla-bla actually has any predictive value. The opposite could be as true!

So what to do? I suggest to always backtest and validate any trading idea. Yes, you’ll still just look at the past and backtesting has its flaws but it surely puts you ahead of a trader who doesn’t backtest at all. If you don’t believe in backtesting, simply ask yourself if you would still trade a chart pattern if you’d know that it lost money every year for the last 10 years in the market and timeframe you plan to trade it. I wouldn’t and that’s why I want to know at least that much before betting my money on an idea.

Backtesting the right way is not easy. You need to learn how to do it as well as possible, knowing it’s limitations and flaws. Do your best to avoid over optimizing, make sure your backtest is statistically sound and you have enough trade samples (no, 30 is not enough). Evaluate trading costs correctly and use high-quality historical data. Depending on the strategy, use out of sample validation, walk-forward optimization, cross validations and monte carlo simulations.

If you’re doing proper backtesting, you’ll be way ahead of all the traders out there who just blindly trade some signal. You’ll know a lot more about the possibly good signals and you’ll have filtered out tons of strategies that you now know for sure have no edge whatsoever.

Even now though you’ll still not know for sure if your strategy really has an edge or not. You’ve done all you can to strongly improve the odds but it still is a speculative edge and should be treated as such.

That’s why I see trading systems as investments. After having done my research, I might decide to invest in a trading system. And like investing in stocks, I’ll diversify as good as possible by investing in multiple systems. This way if some of my systems are actually just “data accidents” that somehow still made it through my strict system development process, it’s not going to be a disaster as long as the other systems perform as expected.

I believe that’s as close as you can come to a sure thing edge as a speculative trader. Of course, that’s not what traders want to hear but it’s always much better to know and be aware of what’s true rather than to ignore it.

PS: If you know of any sure thing trading edges out there, please email me!

Happy Trading!

Marco



subscribe to the algostrats newsletter